Fill an Instacart cart for a risotto recipe with an AI agent: ChatGPT Dots did it, Gemini Spark froze
- Posted
- by Ruben Circelli (PCMag)
- Last checked
- Method:
- How we verify
Evidence: Article or interviewA newsletter, journalist or interview account of what happened.
PartialThe source says it only partly worked, or had a significant problem a person had to deal with.
What went wrong: One agent completed the task; one added an item that wasn't in the recipe and one froze.
Context: In a comparison test: from a head-to-head chatbot review, not a standalone agent test.
Outcome
For PCMag, Ruben Circelli asked three agents to find J. Kenji López-Alt's no-stir risotto recipe, research the best version of each ingredient at the local Wegmans and add them to an Instacart cart. ChatGPT's Dots, once Instacart was connected, found the recipe, researched the ingredients and filled the cart without issue. Meta's Muse got there slowly, needed Circelli to log in and enter a verification code, and added saffron threads that aren't in the recipe. Gemini's Spark found the recipe and proposed ingredients but couldn't access the account, tried to take over Chrome and froze.
Agent
ChatGPT Dots, Muse, Gemini Spark
Prompt Recreated prompt
The source didn't share its exact prompt, so we wrote this from what it describes.
Run this task now, exactly as written below. Everything you need is here, so start straight away and only stop if something required is missing or a rule below tells you to. Goal: Find a recipe and add exactly its ingredients to my online grocery cart at my usual store. Fill in before running: - The recipe (name or link): [RECIPE] - My store and grocery service: [STORE] - How many servings and anything I already have: [DETAILS] Rules: - Add only what the recipe calls for, in the right quantities; nothing extra. - Pick a good-value version of each item and say why. - List anything out of stock and the substitute you'd suggest. - Leave the cart for me to review; don't check out. - If a site asks me to sign in, pause and let me do it myself (including any CAPTCHA or two-step code); never guess or store passwords. - Ask for my OK before you pay, book, send, submit or cancel anything. - Check facts that can change (prices, opening hours, dates, availability, rules) against the official source, and tell me anything you couldn't verify. Output: A filled cart and a list of each item, price and any substitutions. You'll know it worked when: the cart matches the recipe and I only have to press checkout.
Customize this prompt
Answer these questions and we'll put your answers into the prompt for you. You can leave any of them empty.
Your prompt
Run this task now, exactly as written below. Everything you need is here, so start straight away and only stop if something required is missing or a rule below tells you to. Goal: Find a recipe and add exactly its ingredients to my online grocery cart at my usual store. Fill in before running: - The recipe (name or link): [RECIPE] - My store and grocery service: [STORE] - How many servings and anything I already have: [DETAILS] Rules: - Add only what the recipe calls for, in the right quantities; nothing extra. - Pick a good-value version of each item and say why. - List anything out of stock and the substitute you'd suggest. - Leave the cart for me to review; don't check out. - If a site asks me to sign in, pause and let me do it myself (including any CAPTCHA or two-step code); never guess or store passwords. - Ask for my OK before you pay, book, send, submit or cancel anything. - Check facts that can change (prices, opening hours, dates, availability, rules) against the official source, and tell me anything you couldn't verify. Output: A filled cart and a list of each item, price and any substitutions. You'll know it worked when: the cart matches the recipe and I only have to press checkout.
Scroll to see the whole prompt. Your answers show in green.
3 blanks are still empty: The recipe (name or link); My store and grocery service; How many servings and anything I already have. They'll stay in brackets in your prompt, so the chat tool may ask you about them.
What you type stays in this browser tab. We don't send it or save it, and it's gone when you leave the page. If you use a "Try this in" button, your answers go to that tool as part of the prompt.
Try this in:
This prompt has blanks to fill in, like [RECIPE]. We'll copy it instead of filling it in, so you can edit it before you send.
Opens in a new tab. We don't see what you send.
These links are only a shortcut. We're not affiliated with OpenAI, Google or xAI, and your results may not match the source.
Optional stronger version
Our rewrite with extra checks. The poster didn't test it.
Run this task now, exactly as written below. Everything you need is here, so start straight away and only stop if something required is missing or a rule below tells you to. Goal: Find a recipe and add exactly its ingredients to my online grocery cart at my usual store. Fill in before running: - The recipe (name or link): [RECIPE] - My store and grocery service: [STORE] - How many servings and anything I already have: [DETAILS] Rules: - Add only what the recipe calls for, in the right quantities; nothing extra. - Pick a good-value version of each item and say why. - List anything out of stock and the substitute you'd suggest. - Leave the cart for me to review; don't check out. - If a site asks me to sign in, pause and let me do it myself (including any CAPTCHA or two-step code); never guess or store passwords. - Ask for my OK before you pay, book, send, submit or cancel anything. - Check facts that can change (prices, opening hours, dates, availability, rules) against the official source, and tell me anything you couldn't verify. - Show the cart total and flag any item that costs much more than usual. Output: A filled cart and a list of each item, price and any substitutions. You'll know it worked when: the cart matches the recipe and I only have to press checkout. Before you finish: re-read my goal and constraints, check your answer against each one, and list anything you assumed or couldn't confirm.
Source
https://uk.pcmag.com/ai/167960/i-let-3-ai-agents-take-over-my-life-only-1-didnt-fail
Related cases
Sky Canaves tried OpenAI's example 'plan and buy ingredients to make Japanese breakfast for four'; ChatGPT agent shopped her local H-E-B through Instacart, needed her to take over to log in, filled a roughly $130 cart with things she already had, asked before paying and the trimmed order arrived on time. 'It worked, yes', she wrote, but with plenty of friction.
ChatGPT agent · Shopping · Source date:
- Order groceries from a 16-item list with OpenAI OperatorRecreated promptEvidence: Article or interview
Operator misread parts of a 16-item grocery list and added only 13 items to the cart even after corrections; the author concluded it wasn't ready for real use.
OpenAI Operator · Shopping · Source date:
- Fill an online grocery cart from a handwritten list with ChatGPT agentImproved promptEvidence: Article or interviewPartial
ChatGPT agent transcribed 15 of 16 items from a photographed shopping list (missing onions), found the nearest Harris Teeter and added them to the cart; a safety monitor first blocked the Kroger login, but after the user said they'd log in themselves and entered the password, it ordered the groceries.
ChatGPT agent · Shopping · Source date: